This is a web-app which reads data from an accelerometer worn on the body (such as a mobile phone) and uses machine-learning models to classify how physically tired the person is. The goal is to improve workplace safety - it could be used on factory floors, construction sites, etc.

The front-end is a Streamlit app, and the back-end is a voting block of machine-learning models built in Scikit-learn, and a neural network built in Tensorflow, trained on data from published biophysics studies.